[Bug tree-optimization/106243] New: Failure to optimize (0 - x) & 1 on vector type

[Bug tree-optimization/106243] New: Failure to optimize (0 - x) & 1 on vector type

[gabravier at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106243

            Bug ID: 106243
           Summary: Failure to optimize (0 - x) & 1 on vector type
           Product: gcc
           Version: 13.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: gabravier at gmail dot com
  Target Milestone: ---

#include <stdint.h>

typedef int64_t v2i64 __attribute__((vector_size(16)));

v2i64 f(v2i64 x)
{
    return (0 - x) & 1;
}

This can be optimized to `return x & 1;`. LLVM does this transformation, but
GCC does not

[Bug tree-optimization/106243] Failure to optimize (0 - x) & 1 on gimple level (including vector types)

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106243

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|Failure to optimize (0 - x) |Failure to optimize (0 - x)
                   |& 1 on vector type          |& 1 on gimple level
                   |                            |(including vector types)
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2022-07-10
             Status|UNCONFIRMED                 |NEW

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Confirmed, even scalar version is not optimized on the gimple level:
int f1(int x)
{
    x = -x;
    return x & 1;
}


RTL level is done during combine for the scalar version:
Trying 6 -> 7:
    6: {r86:SI=-r87:SI;clobber flags:CC;}
      REG_DEAD r87:SI
      REG_UNUSED flags:CC
    7: {r85:SI=r86:SI&0x1;clobber flags:CC;}
      REG_DEAD r86:SI
      REG_UNUSED flags:CC
Successfully matched this instruction:
(parallel [
        (set (reg:SI 85)
            (and:SI (reg:SI 87)
                (const_int 1 [0x1])))
        (clobber (reg:CC 17 flags))
    ])

[Bug tree-optimization/106243] Failure to optimize (0 - x) & 1 on gimple level (including vector types)

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106243

--- Comment #2 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The trunk branch has been updated by Sam Feifer <sfeifer@gcc.gnu.org>:

https://gcc.gnu.org/g:39579ba8de9d91eafcc4943259b154a9e66538f1

commit r13-1962-g39579ba8de9d91eafcc4943259b154a9e66538f1
Author: Sam Feifer <sfeifer@redhat.com>
Date:   Wed Aug 3 10:31:03 2022 -0400

    match.pd: Add bitwise and pattern [PR106243]

    This patch adds a new optimization to match.pd. The pattern, -x & 1,
    now gets simplified to x & 1, reducing the number of instructions
    produced.

    This patch also adds tests for the optimization rule.

    Bootstrapped/regtested on x86_64-pc-linux-gnu.

            PR tree-optimization/106243

    gcc/ChangeLog:

            * match.pd (-x & 1): New simplification.

    gcc/testsuite/ChangeLog:

            * gcc.dg/pr106243-1.c: New test.
            * gcc.dg/pr106243.c: New test.

[Bug tree-optimization/106243] Failure to optimize (0 - x) & 1 on gimple level (including vector types)

[roger at nextmovesoftware dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106243

Roger Sayle <roger at nextmovesoftware dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |FIXED
                 CC|                            |roger at nextmovesoftware dot com
             Status|NEW                         |RESOLVED
   Target Milestone|---                         |13.0

--- Comment #3 from Roger Sayle <roger at nextmovesoftware dot com> ---
This has now been fixed on mainline (for both scalar and vector types).  Thanks
Sam.

[Bug tree-optimization/106243] Failure to optimize (0 - x) & 1 on gimple level (including vector types)

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106243

--- Comment #4 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Jakub Jelinek <jakub@gcc.gnu.org>:

https://gcc.gnu.org/g:621f5362253f00b910686e8221e6756457f71e81

commit r13-2019-g621f5362253f00b910686e8221e6756457f71e81
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Thu Aug 11 10:29:06 2022 +0200

    testsuite: Fix up pr106243* tests on i686-linux [PR106243]

    These 2 tests were FAILing on i686-linux or e.g. with
    --target_board=unix/-m32/-mno-sse on x86_64-linux due to
    -Wpsabi warnings.

    2022-08-11  Jakub Jelinek  <jakub@redhat.com>

            PR tree-optimization/106243
            * gcc.dg/pr106243.c: Add -Wno-psabi to dg-options.
            * gcc.dg/pr106243-1.c: Likewise.